2.6 Modelling skills
Studying this course? Keep your progress in one place.
Key facts
- Modelling turns a real situation into mathematics and back again.
- Choose the model from the shape of the data, the curve's properties and the context.
- The domain is the set of inputs that make sense in the real situation.
- Find parameters from initial conditions, substitution of points or simultaneous equations on the GDC.
- Test and reflect: is the model appropriate and reasonable?
- Extrapolation (outside the data) can be very unreliable; interpolation is usually reliable.
The modelling process
Modelling is a cycle: develop a model, fit it to the data, test it, then use it.
If the model is not reasonable, return to the first step and choose or adjust it.
You use the SL 2.5 models (linear, quadratic, exponential, direct and inverse variation, cubic, sinusoidal). Fitting a model by regression is topic 4, and non-linear regression is not expected at SL.
- 1
Develop
understand the context, choose a function type
- 2
Fit
find the parameters from the data
- 3
Test and reflect
compare the model with the data and context
- 4
Use
read values, interpret, predict
Then back to step 1
In which step do you decide whether a model gives realistic values?
Choosing a model and domain
Match the shape of the data to a function type, then restrict the domain to what is physically possible.
Choose from the shape of the data, the properties of the curve and the context.
The domain is the set of inputs that make sense. Time cannot be negative, a tank cannot hold more than its capacity and a count cannot be fractional.
Linear
- Constant rate of change
Quadratic
- One maximum or minimum
Exponential
- Same multiplying factor for equal steps
- Levels off at an asymptote
Sinusoidal
- Regular repeating cycle
Worked example
A tank holds 40 litres and is filled at 25 litres per hour, so . Its capacity is 400 litres. Find the domain.
- 1
Time cannot be negative, so .
- 2
The volume cannot exceed capacity: .
- 3
Solve: , so .
A model gives the number of people in a room, . Which domain feature must a sensible model have?
Finding the parameters
You can fix parameters from initial conditions, from known points, or by solving simultaneous equations on the GDC.
Initial conditions: for , . For : , , .
Substitution of points: through and with : , , so .
Simultaneous equations: up to three linear equations in three variables, solved with technology.
- Sinusoidal
- Sinusoidal
- Sinusoidal
Worked example
Find , and for through , and .
- 1
gives .
- 2
gives , so .
- 3
gives , so .
- 4
Subtract: , so and .
A sinusoidal model has maximum 14 and minimum 4. What is in ?
Testing and reflecting
Test the model against the data and the context, and say which assumption fails if it gives an unreasonable value.
Comment on whether the model is appropriate and reasonable:
- Does the curve pass close to the data and have the right shape?
- Are predicted values sensible (no negative lengths or populations, no volume above capacity)?
- Is the domain sensible?
Justify a choice with specific features, for example "the mass halves every 6 days, so an exponential model is appropriate".
Appropriate
- The shape and features match the type of situation
Reasonable
- The values it gives are realistic
A quadratic model for a town's population predicts a negative population in 2100. What should you say?
Using the model
Predicting inside the data range is usually reliable; predicting outside it can be very unreliable.
Use the model to read values, solve equations on the GDC and interpret answers in context with units.
Interpolation is predicting within the range of the data. Extrapolation is predicting outside it. A bacteria model would give impossible numbers for large , because food and space run out.
A model is built from data for to . Which prediction is least reliable?
Try an exam question
A quadratic model was fitted to a town's population (thousands) for . It predicts a population of zero at . Explain whether the model is reasonable for large .
[3 marks]
- [1]The prediction at is far outside the data range (extrapolation).
- [1]A population of zero is unrealistic for a growing town.
- [1]So the model is only appropriate near the data range () and should not be trusted beyond it.
That's the notes covered.
Carry on to the next subtopic.
Try a question
Which of the following is the most reasonable domain for the model?
1 mark, marked against the mark scheme
Which of the following is the most reasonable domain for the model?
1 mark, marked against the mark scheme